如何 select 在大查询中按 ID 分组的前 N 行？

Question

我有一个（已更改 items/data）杂货店过道和产品的数据库，我正试图在每个过道中找到有货的前 3 件最昂贵的商品。

我的大查询代码是这样的：

SELECT ROW_NUMBER() OVER (PARTITION BY aisle ORDER BY price DESC) AS high_prices, aisle, 
price, product FROM `a_database.shopping.grocery_stores`
WHERE in_stock = TRUE AND high_prices <= 3
GROUP BY aisle 
ORDER BY price;

由于 high_prices <= 3，我收到错误：unrecognized name: high_prices at [2:30]，但是，我不知道其他方法。

如前所述，我希望输出商店过道上有库存的 4 件最昂贵的商品。但由于上述错误，它不起作用。当我删除 high_prices <= 3 时，我得到的是商店中所有商品的列表以及它们完全乱序的过道。

如何获得使该查询可行的结果？

预期输出类似于此屏幕截图：

我包含了一个示例数据 table:

aisle    product        in_stock    price
1        fresh_chicken  TRUE        10
2        frozen_fish    TRUE        10
3        truffles       TRUE        15
2        cereal         TRUE        5
2        cake           TRUE        25
3        wine           TRUE        30
1        fresh_fish     TRUE        15
1        fresh_beef     TRUE        10
2        seasonings     FALSE       15
3        whiskey        TRUE        25
1        whiskey        TRUE        40
1        pre_cooked     TRUE        30
2        fresh_pie      FALSE       10
3        fresh_salad    FALSE       5

非常感谢任何帮助或建议。

Answer 1

如果要引用别名，请使用 HAVING

SELECT *  FROM
    (SELECT ROW_NUMBER() OVER (PARTITION BY aisle ORDER BY price DESC) AS high_prices, aisle, 
       price, product 
    FROM `a_database.shopping.grocery_stores`
    WHERE in_stock = TRUE
    GROUP BY aisle ) t1 
WHERE high_prices <= 3
ORDER BY price;

Answer 2

下面使用

select *
from your_table
where in_stock
qualify 3 >= row_number() over(partition by aisle order by price desc)

如果应用于您问题中的示例数据 - 输出为

Answer 3

使用子查询：

select *
from (select row_number() over (partition by aisle order by price desc) row_number,
             aisle,
             price,
             product
      from a_database.shopping.grocery_stores
      where in_stock is true) a
where row_number <= 3

如何 select 在大查询中按 ID 分组的前 N 行？

How do I select top N rows grouped by an ID in big query?

sql

google-bigquery

如何 select 在大查询中按 ID 分组的前 N ​​行？

How do I select top N rows grouped by an ID in big query?

sql

google-bigquery

如何 select 在大查询中按 ID 分组的前 N 行？